DESCRIPTION: (Verbatim from the Applicant's Abstract) Accurate seizure characterization and localization can be important for appropriate treatment of patients with epilepsy including those with intractable partial epilepsy being considered for seizure surgery. While in some patients the regions of seizure onset may be apparent from scalp EEG recordings, not infrequently scap recordings are not sufficient and intra-cranial ictal EEG recordings must be obtained using depth electrode arrays or subdural grids and strips. Some seizures, particularly those originating from frontal lobe regions or other neocortical regions (e.g., temporal lobe) may be particularly difficult to localize even with extensive incracranial recording arrays because of broad regional onset or rapid seizure propagation. New computer assisted methods for EEG analysis can provide significant insights not apparent from visual EEG inspection. The ability to measure patterns of seizure flow, using such methods as the directed transfer function method can determine sources of seizure activity that are not otherwise apparent. The recently described matching pursuit method allows detailed and continuous time-frequency decomposition of entire seizures, allowing exact characterization of the component frequencies. The specific aims of this proposal are to apply these various analytic methods to characterize both mesial temporal onset seizures and seizures originating from lateral temporal and frontal neocortical regions, specifically addressing seizure dynamics and component frequencies. These and other techniques will be used to address the important issues regarding seizure occurrence. The occurrence of an epileptic seizure may not be a random event, particularly wen clusters of seizures occur. This proposal will apply various measures of synchronization and complexity to provide insights into evidence for deterministic behaviors that govern seizure occurrence, recurrence and seizure termination. Understanding the spatiotemporal relationships that produce seizure can provide important insights into epileptogenesis.